Page 7 - Demo

S.K. Singh et al. | 18534.1 Spectral Ratios between Monte Alb%u00e1n and OXLCThe ratios were computed for shallow (H %u2264 35 km) and deep (H > 35km) earthquakes separately. We selected 7 shallow and 5 deep events with good signal to noise ratio for the analysis (Table 3). Fourier spectrum of each component of each event was computed at OXLC and the stations of the MAPN (base, mid-height, and top). The time window chosen for spectral analysis bracketed 95% of the energy in the accelerograms beginning with the arrival of S-wave. A 5% cosine taper was applied before computing the spectrum which was then smoothed by a 1/6 octave filter. The horizontal components were grouped together. The horizontal and vertical ratios along with geometric mean and %u00b1 one standard deviation curves for shallow and deep sources are plotted in Figures 4a and 4b, respectively. We note that: (1) The ratios for shallow and deep events are similar. (2) Horizontal components at the Monte Alb%u00e1n sites are amplified with respect to OXLC in the frequency range 1.5 Hz %u2013 20 Hz; the peak amplification at the base and mid-height is ~ 3 and at the top ~ 10. The peaks occur at f ~ 6 Hz. (3) The amplifications of the vertical components are less than that of the corresponding horizontal components. (4) The ratios show a de-amplification between 0.2 and 1.5 Hz.Theoretical calculations show that topographic high (ridge) causes amplification of seismic waves while depression (canyon) produces de-amplification (e.g., Bouchon, 1973; Boore, 1973). The amplification is high at a wavelength comparable to the ridge dimension. Assuming a shear-wave velocity of 3 km/s, the wave length corresponding to 6 Hz is 500 m, which is close to the difference in the height of Monte Alb%u00e1n and OXLC. De-amplification caused by depression occurs at lower frequencies which is the case at Monte Alb%u00e1n. However, topographic map does not show any depression between OXLC and Monte Alb%u00e1n (Figure 1).4.2 Some TestsHuatulco earthquake of 23/06/2020 (Mw7.4) (Figure 1) was recorded both at OXLC and the triaxial sensor located at midheight of MAPN. These recordings permit us: (1) to examine whether spectral ratio at MAPN with respect to OXLC computed for the 2020 earthquake differs from that obtained using moderate events (a test of possible nonlinearity), (2) to observe similarity and difference in the recorded ground motions at the two sites, and (3) to check the validity of the random vibration technique by comparing the observed and estimated ground motions.Figures 5a and 5b illustrate these motions. We note that velocity and displacement traces at the two sites are very similar in waveform and amplitude; the acceleration at the mid-height of MAPN, however, is amplified, an expected result from Figure4. In other words, at longer periods the motions are the same so that the velocity and displacement traces at OXLC may be used for MAPN.For comparison of observed and estimated motion at midheight during 2020, we select pseudoacceleration response spectrum, Sa, 5% damping. Sa at mid-height was estimated with median spectral ratio for shallower (H %u2264 35 km) events (the depth of the earthquake was 17 km). Effective duration, Te, a parameter needed in the calculations, was taken as the duration of the time window that brackets 95% of the energy in the accelerogram at OXLC. For the 2020 event, Te of NS, EW, and Z components are 16.6, 18.5, and 34.7 s, respectively; the corresponding values at Table 3. Earthquakes recorded both at OXLC and Monte Alb%u00e1n which were used in computing transfer functions (spectral ratios)Date Lat (%u00b0N) Long (%u00b0W) H, km M06/07/2007 16.90 94.10 100 6.212/02/2008 16.19 94.54 90 6.609/02/2010 15.90 96.86 37 5.825/06/2010 15.22 96.44 12 5.107/04/2011 17.2 94.34 167 6.705/05/2011 16.61 98.91 11 5.502/10/2019 15.57 95.06 12 5.323/06/2020 15.78 96.12 22 7.424/06/2020 15.48 96.57 8 5.502/07/2020 16.21 98.02 19 5.224/07/2020 15.36 96.04 15 5.708/12/2020 15.65 95.41 53 5.5